从Jupyter Notebook发布Azure ML Web服务

问题描述

发布Azure ML Web服务时来自Jupyter笔记本，我们遇到了以下错误..

ValueErrorTraceback (most recent call last)
<ipython-input-4-7e2aa3c8d286> in <module>()
      4 @services.publish('id','token')
      5 @services.types(InvoiceNo=str,StockCode=str, Description=str,Quantity=float,UnitPrice=float,CustomerID=int,Country=str)
----> 6 @services.returns(str)
      7
      8

/home/nbuser/anaconda2_20/lib/python2.7/site-packages/azureml/services.pyc in do_publish(func)
    939     if not callable(func_or_workspace_id):
    940         def do_publish(func):
--> 941             func.service = _publish_worker(func, files, func_or_workspace_id, workspace_id_or_token, endpoint)
    942             return func
    943         return do_publish

/home/nbuser/anaconda2_20/lib/python2.7/site-packages/azureml/services.pyc in _publish_worker(func, files, workspace_id, workspace_token, management_endpoint)
    883             msg = str(resp.status_code)
    884         raise ValueError('Failed to publish function: ' + msg + chr(10)  +
--> 885                          'Set azureml.services._DEBUG = True to enable writing {}.req/{}.res files'.format(func.__name__, func.__name__))
    886
    887     j = resp.json()

ValueError: Failed to publish function: The request payload exceeds the maximum allowed request size of 134217728 bytes. Reduce the request payload size and try again.
Set azureml.services._DEBUG = True to enable writing customer_segmentor_Web.req/customer_segmentor_Web.res files

推荐答案

这篇关于从Jupyter Notebook发布Azure ML Web服务的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！